Not lastly, the 1990s saw the rise of post-
structuralist influences in geography. Above all,
poststructuralism was interpreted as an epis-
temological critique, with advocates arguing
that previously sacrosanct ontological cate-
gories lacked foundational status (Dixon and
Jones, 1998; Doel, 1999). Derrida’s (1988)
‘constitutive outside’ has been influential in
this regard. He proposes an anti-foundational
theory of concept construction that rejects
the structuralist positivity essential to ground-
ing philosophical approaches to God, man and
self (Derrida, 1970). Instead, knowledge is only
possible through an exclusionary, negating
process, one that leaves a (deconstructive)
‘trace’ of the other within the inscribed
boundaries of the categories of knowing. This
formulation has been used to destabilize
notions of a fixed subject, replacing an essen-
tial identity with a socially constructed cate-
gory defined by the constitutive outside – the
raw material for the formation of identity
(Laclau and Mouffe, 1985; Natter and Jones,
1997). Geographers have also employed
Foucault’s (1970; 1972) archaeological method
to uncover how clusters of power/knowledge
produce, fill, and maintain categories, such as
nature (Willems-Braun, 1997; also see Philo,
1992). Both sorts of analysis typically dissolve
ontological certainties: the point is no longer
what we know, but how we came to know
what we know in the first place. Such work is
best done from a cross-cultural and historical
perspective, one that admits that how we
describe the world is constrained by the
place- and time-bound languages that we have
at our disposal (see Howitt and Suchet,
Chapter 31 in this section).
In addition to nature–culture and identity,
poststructuralist geographers have also called
into question the concept of scale, which,
rather than being viewed as an ontological cate-
gory derived from a foundational spatiality
(á laNystuen, 1963), can be understood as
both a discourse and the spatial counterpart
to the general–particular epistemological
opposition (Jones, 1998). Nor did the concept
of culture escape the broom-sweep of dis-
course: Mitchell (1995) argues that we should
pay attention not to culture’s attributes –
which he finds chaotically conceived – but to
the work done in its name. Gibson-Graham
(1996) makes parallel arguments for capitalismand other tropes related to the economy.
Other key concepts subject to deconstruction
include representation (Deutsche, 1991;
Harley, 1989; Jones and Natter, 1999) and
space/place (Hooper, 2001).There have, finally,
been a few attempts to theorize a poststruc-
turalist spatial ontology (Massey, 1994; Rose,
1993; Soja, 1996). The most radical account
thus far appears in Doel’s Poststructuralist Geo-
graphies (1999). Methodologically, he rejects
dialectics in favor of deconstruction, and though
he claims ontological agnosticism, he is also
supportive of Deleuze and Guattari’s (1987)
folded, rhizomatic, and flowlike spatiality –
what he terms ‘scrumpled geography’.
In summary, geography’s own knowledge
production has relied upon historically contin-
gent deployments of a handful of key binary
oppositions. Over time, the assumption of an
orderly world has given way to an assumption
of disorder – even within physical geography
(Phillips, 1999). Nature and culture, two con-
cepts – or facts of life, depending on your per-
spective – once set the parameters within
which geography was practiced and organized;
after a period of relative neglect, this opposi-
tion is now one of the most intense areas of
theoretical reflection in the field (see Braun
and Castree, 1998; Castree and Braun, 2001).
The problem of integrating the individual and
society (or agency–structure), long over-
looked and then seemingly solved through
structuration theory during the 1980s, has
re-emerged with the erasure of the self under
poststructuralism’s assault on identity, and
with the arrival of psychoanalytic approaches
(Nast, 2000; Pile, 1996). And though most
geographers still tend to work with an ontolog-
ical division between space and place (Entrikin,
1991) – a legacy of the separation between
scientific and humanistic approaches –
Hooper’s (2001) analysis indicates that this
too might be an epistemological division. Not
lastly, geography continues to be haunted,
along with most other disciplines, by the dis-
cursive versus ‘real world’ division (see Peet,
1998).This offshoot of the idealism–materialism
opposition has created a considerable barrier
between cultural and economic geographers,
despite numerous attempts at integration
(for example, Harvey, 1996; see Barnes, intro-
duction to Section 2 in this volume, for a
discussion).516 SPACES OF KNOWLEDGESection-9.qxd 03-10-02 10:45 AM Page 516